A factor set-based GNSS fault detection and exclusion for vehicle navigation in urban environments

Abstract

With the rapid development of safety critical applications of Intelligent Transportation Systems (ITS), Global Navigation Satellite System (GNSS) fault detection and exclusion (FDE) methods have made navigation systems increasingly reliable.However, in multi-fault scenarios of urban environments, FDE methods generally demand massive calculations and have a high risk of missed detection and false alarm.To deal with this issue, we proposed a factor set based FDE algorithm for the integration of GNSS and Inertial Measurement Units (IMU). The FDE is first performed efficiently via consistency checking over far fewer subsets of the pseudorange. Afterwards, the FDE results are validated by missed-detection and false-alarm checks. The misseddetection-check factor is designed by predicting the maximum horizontal GNSS positioning error, while the false-alarm-check factor is designed with the aid of IMU mechanization. Following FDE, a loosely coupled GNSS/IMU integration is carried out to output the final estimation of the position, velocity and attitude of the vehicle.The proposed algorithm improved both horizontal and 3D positioning accuracy by more than 50% in the field test, compared to the traditional GNSS/IMU loosely coupled scheme. Additionally, with the proposed algorithm, the resultant accuracy of the velocity and of the heading angle were improved by over 20% and 50% respectively.

Carrier Phase Residual Modeling and Fault Monitoring Using Short-Baseline Double Difference and Machine Learning

“…We chose the larger of the two values to obtain the conservative inflation factors, which were 8.31 for HPF and 7.82 for LPF. Finally, the inflation fac- In this study, we demonstrate the computation of the inflation factor and resulting threshold to meet a false alarm requirement of 10 −5 , which has been used in various previous works [41,52,53]. We computed the terms from the training dataset and then applied them to the test dataset to verify the false alarm probability.…”

A computationally efficient Kalman filter-based RAIM algorithm for aircraft navigation with GPS and NavIC

“…In order to compute the PL, the maximum FMS corresponding to α 2 fe needs to be calculated for the effect of worst-case faults. For this purpose, fault contributions of each epoch's innovation term are required using equation (21) to form the necessary coefficient matrix in the denominator of the FMS. Let f fe, ℓ be the fault vector at time ℓ, and F full, ℓ−1 and K fe, ℓ be the state transition matrix and Kalman gain, respectively, of the full-order EKF.…”

“…In order to avoid the architectural complexity with multiple parallel filters, KF integrity monitors in the range domain have been explored, where fault detection tests are designed with measurement innovations/residuals [15][16][17][18][19][20][21]. This method is called range-based integrity monitoring in this paper.…”

“…Performance of the presented method is studied for different GNSS fault types and IMU grades. A factor set-based fault detection and exclusion (FDE) strategy is presented in [21] for loosely coupled GNSS-INS integration. The FDE works effectively with a few subsets of pseudorange measurements.…”

ConvGRU-MHM: a CNN GRU-enhanced MHM for mitigating GNSS multipath

“…Improvements in the GNSS mathematical models using artificial intelligence have significant advantages [1]. In fields such as autonomous vehicles, indoor and outdoor positioning, and intelligent transportation systems [2][3][4], artificial intelligence techniques have been used to improve the completeness, continuity, robustness, and accuracy of mathematical models for navigation and positioning. Before being combined with artificial intelligence, GNSS mathematical models (functional and stochastic) were used to achieve an optimal estimation of GNSS parameters [5,6].…”